1. Maria Grazia Pia, INFN Genova 1 New models for PIXE simulation with Geant4 CHEP 2009 Prague, 21-27 March 2009 Maria Grazia Pia INFN Genova G. Weidenspointner, M. Augelli, L. Quintieri, P. Saracco, M. Sudhakar, A. Zoglauer INFN Sezione di Genova and INFN Laboratori Nazionali di Frascati, Italy Space Sciences Laboratory, UC Berkeley, USA CNES, Toulouse, France Max-Planck-Institut für extraterrestrische Physik and Halbleiterlabor, Germany

2. Maria Grazia Pia, INFN Genova 2 PIXE P article I nduced X -ray E mission Particle = proton, ion (electron) The physics of PIXE is simple The simulation of PIXE is NOT simple

3. Maria Grazia Pia, INFN Genova 3 PIXE simulation recipe Create a vacancy in the shell occupancy due to ionisation De-excite the atom through fluorescence and Auger emission Proton (ion) ionisation cross sections for individual shells Radiative and non-radiative transition probabilities Atomic binding energies (X-ray energies) Ingredients Instructions Infrared divergence Infrared divergence of the cross section for  -ray emission A touch of spice Monte Carlo codes neutralize the spice by introducing “cuts” (i.e. a threshold for  -ray production) Below the threshold: ionisation treated as continuous energy loss Above the threshold: emission of an electron PIXE is and intrinsically discrete process, which is intimately associated with a continuous-discrete one One of the ingredients (ionisation cross sections) is available in limited quantity (for a few inner shells only) Note

4. Maria Grazia Pia, INFN Genova 4 PIXE in Geant4 (and other codes) Specialized codes for PIXE analysis (and simulation) GeoPIXE, GUPIX, PIXAN, PixeKLM, Sapix,WinAxil, Wits-HEX etc. Only concerned with experimental observables relevant to material analysis Impact ionisation by electrons EGS*, Penelope Cross sections available for all shells “independent” discrete process + “regular” continuous-discrete ionisation Penelope literature reports cases of negative energy deposits in volumes subject to large fluctuations Geant4 struggle with PIXE started in 2001 It is still an open issue, that motivated a new R&D project Not just on PIXE, but on fundamental methods of particle transport…

5. Maria Grazia Pia, INFN Genova 5 PIXE history in Geant4 Geant4 Low Energy Electromagnetic package No PIXE in the first implementation of hadron/ion ionisation total ionisation cross section as in Phys. Ref. Manual (M. Maire & L. Urban, Geant4 Standard EM) PIXE embedded in hadron ionisation process Design to handle multiple shell cross section models (MGP) Shell ionisation cross section models used by G4hLowEnergyIonisation Gryzinski model Shell ionisation cross sections, 1  Z  92, all shells as in EADL Implementation: S. Dussoni, V. Ivanchenko, A. Mantero, S. Saliceti, 2001-2004 Fit to H. Paul’s K-shell experimental database Shell ionisation cross sections, 4  Z  92 Implementation: A. Mantero, S. Saliceti, 2004 Refined fit to H. Paul’s K-shell experimental database Shell ionisation cross sections, 4  Z  92 Implementation: A. Mantero, S. Saliceti, 2004-2005 Shell ionisation cross section models, not used by any process ECPSSR (K), semi-empirical (L), interpolation of H. Paul’s data (K-shell) Implementation: H. Ben Abdelouahed, Geant4 9.2, 2008 NIM B paper H. Abdelhouahed, S. Incerti, A. Mantero, Jan 2009 Geant4 R&D project, THIS TALK… and a 40 page paper Part of a R&D project to address fundamental concepts in particle transport PostStepDoIt + undocumented model in AlongStepDoIt

6. Maria Grazia Pia, INFN Genova 6 First Geant4 PIXE development cycle K shell cross section depends on  -ray production threshold! Depending on the  -ray production cut, the “total” cross section may be smaller than the K shell ionisation cross section Outer shells are neglected Si ionisation by protons Wrong: theory or software or both? R&D ECPSSR (ISICS) + implementations based on fits to empirical database (correct!)

7. Maria Grazia Pia, INFN Genova 7 Pixe in Geant4 9.2 1 st development cycle ECPSSR theoretical (K, p-  ) Paul&Sacher, Paul&Bolik empirical (K, p-  ) Orlic et al. semi-empirical (L, p) R&D models ECPSSR  p Z=6 Z=29 Z=48 Z=79 ECPSSR p  empirical 1 st cycle Z=29 Geant4 9.2 

8. Maria Grazia Pia, INFN Genova 8 Geant4 9.2 performance Complex theoretical calculations (6-8 times slower) 2 orders of magnitude slower than R&D models ~50 times slower than previous Geant4 model for  with same functionality …and much more

9. Maria Grazia Pia, INFN Genova 9 Part of a larger R&D study on transport methods in simulation See talk in Core Tools session on Thursday at 2 pm Still a long way to go… Up to where can one go with existing instruments? Can we deal with PIXE correctly? PIXE is a physically interesting playground …but the issue is the interplay of continuous and discrete transport methods Nanotechnology detectors Radiation effects to components LHC RADMON Short term R&D Long term R&D Gaseous detectors Do we know how to deal with PIXE correctly?

10. Maria Grazia Pia, INFN Genova 10 Crisp domain decomposition No dependence of PIXE component on  -ray production threshold Dependency on “cut” only in client ionisation process Reuse in other context Data-driven model Performance! Use Geant4 Atomic Relaxation component

11. Maria Grazia Pia, INFN Genova 11 Ionisation cross sections Theoretical Empirical To be released as a data library Thanks to RSICC! Subject to rigorous experimental validation Theoretical cross sections tabulated by ISICS thanks to S. Cipolla (thanks to S. Cipolla!)

12. Maria Grazia Pia, INFN Genova 12 Similarities and differences, K K shell cross sections, p Z=6 Z=29 Z=48 Z=79 Kahoul et al. Paul&Sacher ECPSSR ECPSSR Hartree-Slater ECPSSR High E PWBA ECPSSR United Atom

13. Maria Grazia Pia, INFN Genova 13 Similarities and differences, L L shell cross sections, p Z=55 Miyagawa et al. Orlic et al. Sow et al. ECPSSR L1L1 L2L2 L3L3 No systematic, quantitative analysis of all the existing theoretical and empirical models documented in literature

14. Maria Grazia Pia, INFN Genova 14 Experimental validation, K shell Ta F Experimental data from Paul & Sacher compilation Ionisation cross sections X-ray production cross sections

15. Maria Grazia Pia, INFN Genova 15  2 test: p-values K shell, protons

16. Maria Grazia Pia, INFN Genova 16 Categorical analysis Contingency tables Fisher test  2 with Yates correction Pearson  2 (where applicable) Results ECPSSR with Hartree-Slater correction best for K shell cross sections at low energy ECPSSR OK at intermediate energy High energy: scarce measurements based on objective, quantitative analysis 95% CL for not rejecting the null hypothesis

17. Maria Grazia Pia, INFN Genova 17 Same validation process for L shell L 2 Z=79 Experimental data Compilations by Sokhi & Crumpton Orlic, Sow & Tang L 3 Z=47 L tot Z=14 Z=17 L 1 Z=67 ECPSSR

18. Maria Grazia Pia, INFN Genova 18  2 test: p-values Contingency tables

19. Maria Grazia Pia, INFN Genova 19 Results for L shell ECPSSR - United Atom best compatibility with experimental data ECPSSR OK Orlic et al.: worst compatibility with experimental data

20. Maria Grazia Pia, INFN Genova 20 M shell The same analysis will be performed once there are any experimental data…

21. Maria Grazia Pia, INFN Genova 21 Geant4 atomic relaxation M.G. Pia and M. Sudhakar Evaluation of radiative emission rates calculations under review 2007 Geant4 simulation based on EADL Validation of X-ray energies against Des Lattes et al. experimental review compilation Thorough study of theoretical models of emission rates against experimental data

22. Maria Grazia Pia, INFN Genova 22 So, are all the problems solved? Determination of a vacancy Relative probability of ionising a shell w.r.t. the other shells But we do not know the total ionisation cross section (without introducing a dependency on cuts) We can calculate ionisation cross sections only for K, L, M shells OK for light elements Overestimate inner shell vacancies for heavier elements Known, controllable limitation In principle, the issue would be solved if we could calculate ionisation cross sections for all shells

23. Maria Grazia Pia, INFN Genova 23 So, would all the problems be solved? PIXE (i.e. creation of a vacancy) is still associated with the discrete treatment of continuous-discrete ionisation Indirect dependence on cuts Ionisation cross section calculation for outer shells No existing theoretical calculations (except a PWBA attempt on N shell) Hardly any experimental data Long and complex calculations, with assumptions and approximations  -ray emission and ionised shell are not related But there is still a long way to go… We are in better shape than with previous Geant4 The current R&D models can be used for real-life experimental studies real-life experimental studies (knowing their limitations)

24. Maria Grazia Pia, INFN Genova 24 Experimental application Cu shield Cu+Al shield Cu+Al+B 4 C shield Optimisation of the shielding of the X-ray detectors eROSITA telescope on the Spectrum-X-gamma space mission The R&D software is adequate for experimental applications concerned with relative effects of experimentally relevant spectra But absolute calculations are still out of reach for heavier elements (involving outer shells than M) 1 st PIXE spectra ever produced with Geant4!

25. Maria Grazia Pia, INFN Genova 25 R&D on co-working transport schemes in Geant4 Project launched at INFN (2009), international-multidisciplinary team R&D = research study, exploration of novel ideas Distinct from Geant4 production service: no perturbation to running experiments! R&D deliverable(s) = prototypes [to be evaluated for transition into Geant4 releases] Scientific motivation From concrete experimental use cases Objective Seamless transition of simulation régime in Geant4 Capability of simulating complex multi-scale systems Conceptual and software design challenges Physics process adaptation to environment Embedding “mutability” in Monte Carlo physics entities Difficult …not yet present in any simulation system NANO5 Would all the problems be solved? We don’t know it yet…

26. Maria Grazia Pia, INFN Genova 26